Single-cell assignment using multiple-adversarial domain adaptation network with large-scale references.

The rapid accumulation of single-cell RNA-seq data has provided rich resources to characterize various human cell populations. However, achieving accurate cell-type annotation using public references presents challenges due to inconsistent annotations, batch effects, and rare cell types. Here, we introduce SELINA (single-cell identity navigator), an integrative and automatic cell-type annotation framework based on a pre-curated reference atlas spanning various tissues. SELINA employs a multiple-adversarial domain adaptation network to remove batch effects within the reference dataset. Additionally, it enhances the annotation of less frequent cell types by synthetic minority oversampling and fits query data with the reference data using an autoencoder. SELINA culminates in the creation of a comprehensive and uniform reference atlas, encompassing 1.7 million cells covering 230 distinct human cell types. We substantiate its robustness and superiority across a multitude of human tissues. Notably, SELINA could accurately annotate cells within diverse disease contexts. SELINA provides a complete solution for human single-cell RNA-seq data annotation with both python and R packages.

HUSCH: an integrated single-cell transcriptome atlas for human tissue gene expression visualization and analyses.

Understanding gene expression patterns across different human cell types is crucial for investigating mechanisms of cell type differentiation, disease occurrence and progression. The recent development of single-cell RNA-seq (scRNA-seq) technologies significantly boosted the characterization of cell type heterogeneities in different human tissues. However, the huge number of datasets in the public domain also posed challenges in data integration and reuse. We present Human Universal Single Cell Hub (HUSCH, http://husch.comp-genomics.org), an atlas-scale curated database that integrates single-cell transcriptomic profiles of nearly 3 million cells from 185 high-quality human scRNA-seq datasets from 45 different tissues. All the data in HUSCH were uniformly processed and annotated with a standard workflow. In the single dataset module, HUSCH provides interactive gene expression visualization, differentially expressed genes, functional analyses, transcription regulators and cell-cell interaction analyses for each cell type cluster. Besides, HUSCH integrated different datasets in the single tissue module and performs data integration, batch correction, and cell type harmonization. This allows a comprehensive visualization and analysis of gene expression within each tissue based on single-cell datasets from multiple sources and platforms. HUSCH is a flexible and comprehensive data portal that enables searching, visualizing, analyzing, and downloading single-cell gene expression for the human tissue atlas.